Our laboratory studies the molecular pathogenesis of human lymphoid malignancies and has three primary goals: to establish a new molecular diagnosis of human lymphoid malignancies using gene expression profiling, to elucidate the oncogenic pathways that result in malignant transformation of normal B lymphocytes, and to identify molecular targets for development of novel therapeutics for these cancers. To provide a molecular basis for the diagnosis of human lymphoid malignancies, we are exploiting DNA microarray technology to profile gene expression in these cancers on a genomic scale. The laboratory created a novel DNA microarray, the "Lymphochip", which is enriched in genes that are expressed in and/or function in lymphocytes (1). We have used Lymphochip and Affymetrix microarrays to profile gene expression in diffuse large B cell lymphoma (DLBCL) (2-4), chronic lymphocytic leukemia (CLL) (5, 6), mantle cell lymphoma (7), follicular lymphoma (8), multiple myeloma (9), and in a wide variety of normal lymphoid subsets (2, 10-13). One central goal of these studies is to relate gene expression to clinical outcome, thereby establishing a quantitative, reproducible and informative molecular diagnosis of the lymphoid malignancies (14). Our studies have revealed previously unknown types of diffuse large B cell lymphoma that are indistinguishable by current diagnostic methods, but which have strikingly distinct gene expression profiles, originate from different stages of B cell differentiation, utilize distinct oncogenic mechanisms, and differ in their ability to be cured by current chemotherapy (2-4). For several lymphoid malignancies, we have identified molecular profiles that predict the length of survival or the ability to be cured by chemotherapy, thereby providing clinically useful prognostic indicators. Our laboratory has mounted a major effort to create a diagnostic microarray that could provide these molecular diagnoses and prognoses to patients with lymphoid malignancies. Importantly, the genes that are associated with clinical prognosis have provided new targets for therapy of the lymphoid malignancies. Our laboratory uses functional genomics, chemical genetics and molecular biological techniques to validate these and other molecular targets, towards the ultimate goal of targeted therapies for patients aimed directly at the disordered regulatory biology of their individual tumors. MOLECULAR DIAGNOSIS OF LYMPHOID MALIGNANCIES Molecularly and clinically distinct diseases within diffuse large B cell lymphoma (DLBCL) DLBCL has long been enigmatic in that 40 percent of patients can be cured by combination chemotherapy whereas the remainder succumb to this disease. By gene expression profiling, the laboratory discovered that DLBCL is actually comprised of at least three different diseases that are indistinguishable by current diagnostic methods (2-4, 15). As detailed below, these DLBCL subgroups can be considered distinct diseases in that they originate from B cells at different stages of differentiation, utilize distinct oncogenic mechanisms, and differ significantly in their survival rates following chemotherapy. DLBCL subgroups originate from distinct stages of B cell development One subgroup of DLBCL, termed germinal center B cell-like (GCB) DLBCL, expresses genes that are hallmarks of normal germinal center B cells. By contrast, another DLBCL subgroup, termed activated B cell-like (ABC) DLBCL, lacks expression of germinal center B cell-restricted genes and instead expresses genes that are induced during mitogenic stimulation of blood B cells (2). These two subgroups of DLBCL differ in the expression of thousands of genes, and in this respect they are as different as acute myelogenous leukemia is from acute lymphoblastic leukemia. Clues to the normal cellular counterparts of these DLBCL subgroups have been provided by our laboratory's analysis of regulatory factors that control the differentiation of germinal center B cells to plasma cells. We and others showed that BCL-6 is a transcriptional repressor that is required for mature B cells to differentiate into germinal center B cells during an immune response (16, 17). Normal germinal center B cells express BCL-6 at high levels but BCL-6 expression is silenced during plasmacytic differentiation. DLBCLs belonging to the GCB subgroup express BCL-6 at high levels but those belonging to the ABC subgroup do not. BCL-6 is deregulated by chromosomal translocations in roughly 20% of DLBCLs, but the high expression of BCL-6 in GCB DLBCLs is not accounted for by these translocations. Rather, BCL-6 is expressed in GCB DLBCLs along with a host of other germinal center B cell restricted-genes because these DLBCLs are derived from normal germinal center B cells and retain much of their biology. In keeping with this notion, GCB DLBCLs have ongoing somatic hypermutation of their immunoglobulin genes, a characteristic feature of normal germinal center B cells (18). The cell of origin of ABC DLBCL has not been fully elucidated, but may be a plasmablastic B cell that is poised to exit the germinal center. Support for this notion comes from our laboratory's analysis of two regulatory factors that are required for plasmacytic differentiation, Blimp-1 (19) and XBP1 (20). By gene expression profiling, our laboratory demonstrated that BCL-6 blocks the expression of Blimp-1, and when BCL-6 activity was inhibited in a lymphoma cell line, Blimp-1 was induced and plasmacytic differentiation was initiated (21). We went on to show that Blimp-1 is a transcriptional repressor that extinguishes the expression of virtually all germinal center B cell genes, including BCL-6 (22). Blimp-1 and BCL-6 thus form a double negative autoregulatory loop that controls plasmacytic differentiation. Blimp-1 enables the expression of XBP1, which our laboratory showed is a master regulator of the secretory phenotype of plasma cells (23). XBP1 induces the expression of a large set of genes encoding components of the endoplasmic reticulum and golgi, leading to a dramatic expansion of the secretory apparatus (23). In addition, XBP1 increases the overall rate of protein synthesis by 50%, which contributes to the high rate of immunoglobulin secretion by plasma cells (23). In comparison to GCB DLBCLs, ABC DLBCLs are characterized by high expression of XBP1 and its target genes, as well as Blimp-1 (4). This phenotype is similar to that of a rare subpopulation of plasmablasts in the germinal center, which are thought to be in the process of migrating to the bone marrow where they will differentiate fully into plasma cells (24, 25). ABC DLBCLs do not express a variety of other genes that characterize normal plasma cells and multiple myeloma, suggesting that they are derived from a cell that is intermediate between a germinal center B cell and a plasma cell. In support of this notion, ABC DLBCLs have somatically mutated immunoglobulin genes, and therefore are derived from a B cell that has likely traversed the germinal center (18). However, in contrast to GCB DLBCLs, ABC DLBCLs have a fixed complement of immunoglobulin gene mutations, suggesting that the somatic hypermutation machinery has been inactivated as occurs normally during plasmacytic differentiation. Primary mediastinal B cell lymphoma: A distinct subgroup of DLBCL related to Hodgkin lymphoma Recently, we and others developed a molecular diagnosis of a third subgroup of DLBCL, termed primary mediastinal B cell lymphoma (PMBL) (15, 26). PMBL cannot be reliably distinguished from other types of DLBCL by current clinical criteria. PMBL was readily distinguished from GCB and ABC DLBCL by the expression of hundreds of genes, and we were able to develop a molecular diagnosis of PMBL.